143         const v128_t vi2x2 = wasm_v128_load(i2); i2 += 4;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()  local
150 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk00c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
163 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk20c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
171 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk00c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
184 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk20c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
192 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk01c0, wasm_v32x4_shuffle(vi2x2, vi2x2, 2, 2, 2, 2))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
205 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk21c0, wasm_v32x4_shuffle(vi2x2, vi2x2, 2, 2, 2, 2))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
213 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk01c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 3, 3, 3, 3))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
226 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk21c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 3, 3, 3, 3))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
281 …vo1x0 = wasm_f32x4_add(vo1x0, wasm_f32x4_mul(vk02c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
294 …vo0x0 = wasm_f32x4_add(vo0x0, wasm_f32x4_mul(vk22c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
302 …vo1x0 = wasm_f32x4_add(vo1x0, wasm_f32x4_mul(vk02c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
315 …vo0x0 = wasm_f32x4_add(vo0x0, wasm_f32x4_mul(vk22c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
396         v128_t vi2x2 = vzero;  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()  local
403           vi2x2 = wasm_v128_load(i2 + 4);  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
411 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk00c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
424 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk20c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
432 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk00c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
445 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk20c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
454 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk01c0, wasm_v32x4_shuffle(vi2x2, vi2x2, 2, 2, 2, 2))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
471 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk21c0, wasm_v32x4_shuffle(vi2x2, vi2x2, 2, 2, 2, 2))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
481 …vo1x1 = wasm_f32x4_add(vo1x1, wasm_f32x4_mul(vk01c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 3, 3, 3, 3))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
498 …vo0x1 = wasm_f32x4_add(vo0x1, wasm_f32x4_mul(vk21c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 3, 3, 3, 3))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
556 …vo1x0 = wasm_f32x4_add(vo1x0, wasm_f32x4_mul(vk02c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
565 …vo0x0 = wasm_f32x4_add(vo0x0, wasm_f32x4_mul(vk22c1, wasm_v32x4_shuffle(vi2x2, vi2x2, 0, 0, 0, 0))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
571 …vo1x0 = wasm_f32x4_add(vo1x0, wasm_f32x4_mul(vk02c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()
580 …vo0x0 = wasm_f32x4_add(vo0x0, wasm_f32x4_mul(vk22c2, wasm_v32x4_shuffle(vi2x2, vi2x2, 1, 1, 1, 1))…  in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2()